water bubbly flow
 

Hi friends
I need your help! :confused:
my project is about simulation a bubbly flow that has water as primary phase and air as secondary phase!
water injected from top and mixes with air then a column flow with bubbles formed! (bubbles diameter is about 0.3 mm) please see attached pic

In your opinion, is FLUENT able to solve this problem? :rolleyes:
And if yes, which model could help me!? multiphase or DPM!?
any tutorial do you have about liquid flow with air bubbles!?

any idea friends?
any reply could help me...

This seems like a rather tricky setup to me, as your continuous phase switches between the 2 sections. I don't think DPM will do the trick, since your water phase in the bottom will still consist of separate entities in that case. But I don't really see an easy way with Eulerian either - in that case, water will still be the continuous phase in the spray domain, so it will be modeled as a foam there I guess. Is it necesarry to capture the entire domain in a single simulation? Can't you separate it in 2 problems - 1 for the spray part with gas continuous, and 1 for the liquid part with water continuous?

Best,
Cees

Tnx dear cees for your reply
1. A new pic I'm attached! In my problem water injected by an orifice from top and mix with air. Then move down and collected in the bottom and discharged from outlets!
2. So do you think dpm model couldn't helps me?
3. What's your meaning about seperate it in 2 problems? Can you explain more dear friend?

Regarding the DPM part; it would work if it was a solid. but to approximate the continuous water phase as if it were particles, I don't think that would work (unless you move to Lattice-Boltzmann approaches). Maybe the spray behavior would be reasonable, but I have my doubts about the bottom part. I've never done such simulations however, so maybe I'm incorrect - I'd love to hear other ideas on this. (why do you think it might work?)

Regarding the splitting part: I was wondering if you could split the simulation in 2 parts: one modeling what happens above the water surface (so gas continuous, considering only the spray with a uniform gas inflow from the bottom), and one below the water surface (considering gas behavior, and using a UDF to describe water inflow at the top based on the outcomes of simulation 1). I've tried to clarify it in the image below:

I have no idea if this is possible, but it may solve the problem of a change in continuous phase. But, maybe there are solutions to do it in 1 eulerian simulation - I don't know about that

Anyway, why do you want to do this simulation? What do you want to know about the process? Maybe it is not necessary to simulate the whole thing..

https://www.dropbox.com/s/purfk4lxpd...ample.png?dl=0

Thank you very much deer Cees for your reply! :)

1. Does DPM model is suitable only in presence of a solid phase?
2. I'm not familiar with the Lattice-Boltzmann approach! do you think this model could help me?
3. Prediction of behavior of the fluids in the bottom part currently isn't important! Now simulation accurately of the bubble column is only thing that is important for me!
4. I can not see the photo that you attached! Can you attach it somewhere else?
5. finally, I want to model this problem, because I and my supervisor have created such a device and want to calculate the relative velocity of air to water

Thanks for your friendly help

DPM does also work for liquids and gases, but (I think - I've never tested it or so) only if the DPM phase is the discontinuous phase. So, if you are simulating a spray, you can use DPM to simulate water droplets. But the problem is, in the collector, there are no individual water droplets anymore - there is a bulk of water, with gas bubbles rising through it. I don't know how well the DPM model will behave there, but since there is a bulk of water rather than individual drops, I don't think it will do too well.

Regarding lattice-boltzmann, it is a completely separate technique from finite volume, so not something you do with FLUENT. I have no experience in it, I just know it exists, and is based around discretizing the fluid in discrete packages, and moving this packages over a well-structured grid. This works very fast compared to finite volume in some applications (for example, direct numerical simulation of stirred-single phase tanks), but is very ill-developed in others (including chemical reaction and heat transfer for example). I'm not sure about the state of development for multiphase flows. In any case, although some software packages start to become available, it is much more in the state of writing your own software, with research focusing on extending the method to more complex systems. Not something you may want to submerge yourself in if your interest is simulating complex physics with well-established methods.

About 3, I'm a bit confused: the top part is the sprayer and the bottom part the bubble column right? So I guess you are mostly interested in the bottom part?

About the photo, i'll try again.

Best,
Cees

http://i60.tinypic.com/65zbbt.png

Tnx my friend!

you're right! the bubble column is important for me!
I'm attached a photo from one of my simulations! the water inlet behaves like an injector as assumes in the pic!
and I think Eulerian model could solves my problem! Is that wrong?